Tuberculosis (TB) is one of those infections with high morbidity and mortality in all around the world. Hundreds of people died from this disease without diagnosed or due to resistant strains in Turkey. Therefore, it is important to identify postmortem cases who have died from tuberculosis. Molecular methods have been widely used as well as conventional methods in the diagnosis of tuberculosis. The aim of this study was to compare the two different real-time polymerase chain reaction (Rt-PCR) system in the postmortem diagnosis of Mycobacterium tuberculosis infections in paraffin-embedded tissues. A total of 40 paraffin-embedded tissue samples [lung (n= 35), brain (n= 2), heart (n= 2), lymph node (n= 1)] in which histopathologic findings consistent with TB (necrotizing granulomatous inflammation, gelatinous caseous pneumonia, necrotic fibrous nodul) obtained from 37 autopsy cases (31 male, 6 female; age range: 25-85 yrs) were included in the study. Paraffin-embedded tissues were deparafinized with xylene and ethyl alcohol and then DNA isolation was done with QIAsymphony DSP Virus/Pathogen Midi kit in the QIAsymphony device. DNA amplification process was performed by Rt-PCR using the kit Artus® M. tuberculosis RG-PCR in the Rotor-Gene® Q device (Qiagen, Germany). Likewise, after deparafinization process, samples placed in the cartridge and isolation and Rt-PCR was performed by Xpert® MTB/RIF (Cepheid, USA) system, simultaneosly. Seventeen and 20 out of the 40 paraffin-embedded tissues yielded positive results with Qiagen and Xpert system, respectively. M.tuberculosis DNA was found positive in 13 (32.5%) and negative in 16 (40%) of the samples by both of the systems, exhibiting 72.5% (29/40) of concordance. On the other hand, seven (17.5%) samples that were positive with Xpert system yielded negative result with the Qiagen, while four (10%) samples that were positive with Qiagen yielded negative result with the Xpert system. Of the 20 positive cases detected with Xpert MTB/RIF system, 15 were found rifampicin-susceptible, and three were rifampicin-resistant. In two samples in which M. tuberculosis DNA was low positive, rifampicin resistance could not be detected. The identification of M.tuberculosis infections in postmortem cases will contribute epidemiological data in Turkey. In these cases, effective sampling and diagnosing of M.tuberculosis infections by acid-fast stain and culture methods are crucial. However, in cases without microbiological sampling the detection of M.tuberculosis DNA in paraffin-embedded tissues with PCR, although there are differences between PCR systems has diagnostic value. In conclusion, our data indicated that Xpert MTB/RIF system is more favourable to detect M.tuberculosis DNA in paraffin-embedded tissues, with the advantages of determination of rifampicin resistance, and detection of more positive results within a shorter time.